Frontiers in Energy Research (Dec 2024)

State of charge estimation method for lithium-ion battery pack using BP-MDM-TEM

  • Haiyan Yao,
  • Yuefei Xu,
  • Qiang Guo,
  • Yufeng Miu,
  • Xufeng Zhang,
  • Jiadong Wei,
  • Yunning Zhang,
  • Chunshi Wang

DOI
https://doi.org/10.3389/fenrg.2024.1494541
Journal volume & issue
Vol. 12

Abstract

Read online

Addressing the fluctuating core temperatures during series battery pack operation that cause inconsistency in pack battery characteristics, the paper introduces a Battery-Pack Mean Difference-Method With Thermal-Electrical Coupled Model (BP-MDM-TEM). Thermal-electric coupling model describes the mean characteristics of battery pack, while the difference model quantifies variations among individual cells. A dual-time-scale method is employed to estimate State of Charge (SoC) of battery pack, reducing the computational load of the difference model.If the unique battery characteristics exceed system limits, a significant battery method is used to estimate SoC of pack battery. Building upon the BP-MDM-TEM framework,the paper employs the Improved Sage-Husa Adaptive Extended Kalman Filter (ISH-AEKF) to estimate SoC, mitigating the impact of process and measurement noise on system estimation. A hardware test platform is established in this paper, with a lithium-ion battery pack as the research subject.Compared with the traditional MDM method, the Mean Absolute Error (MAE) of soc estimation of BP-MDM-TEM is improved from 0.992% to 0.468%, and the Root Mean Square Error (RMSE) is improved from 1.279% to 0.982%.

Keywords